The main panel of the TwelveKeys screen is where the frequencies of the recording you are analysing are graphed. The horizontal axis of the graph represents pitch, and the vertical axis represents time. Each mark drawn on the screen represents a frequency present in the recording at that particular point in time. So a long, vertical line is likely to represent a sustained note in the recording. The colour and intensity of the marks represent the volume. Loud notes are brighter whereas quieter notes are more faint.
Determining which notes are being played is as simple as lining up the marks with the keys on the piano keyboard at the bottom of the screen. Alternatively you can move your mouse over the marks and the display in the bottom panel will indicate what note you are hovering over. Gridlines can optionally be displayed by right-clicking on the display and selecting "Show Note Lines", or enabling the feature from the options dialog.
When you press play you will hear the recording being played as well as see the graph scroll down the screen. As the marks pass the front edge of the keyboard you will hear them played in the recording. You can think of the front edge of the keyboard as the cursor of the graph. You can press on the keyboard as the notes pass to get a feel for how the graph works. To accurately measure the timing of notes on the display, hover the mouse over the time axis to show a cursor and timestamp.